Get a Demo
See AICap engage a real shopper interaction, live.
Get a DemoProduct — AICap
No hardware. Just a QR code and a question.
AICap is an in-store AI shopping expert. A shopper scans a QR code on the shelf, asks a question the way they'd ask a knowledgeable friend, and gets 2–4 real products from that store's own shelf — explained, compared, and located, in their own language.
Overview
The shopper's own phone is the device.
No app to download, no account to create, no store hardware to buy. A shelf tag and a phone already in a shopper's pocket is the whole system.
Shelf signageA QR shelf strip or bottle tag, in a real aisle, close-up
Shopper scanningA shopper's phone mid-scan, shelf visible in the background
On a store's tabletOptional: AICap running on a tablet the store already owns
- QR code on shelf signage — scans open instantly, no app or signup
- Text or voice — both full experiences, not a fallback for the other
- English, Mandarin, Japanese, and Korean at launch
- Optional tablet or kiosk for stores that want a bigger aisle presence
Getting Started
Sign up in minutes. Live within 48 hours.
No electricians, no hardware to buy, no IT project. One short form and one click to connect your catalog is the whole setup.
- One click grants read-only access to your catalog — Shopify today, leading POS systems next
- Onki verifies and enriches your catalog before you go live
- A print-ready QR signage kit ships with your account — placement tested with you in week one
Setup walkthrough clipShort clip (30–60s): sign up, connect catalog, unbox the signage kit, first scan — for a retailer ops audience
How It Works
From a scan to a confident decision.
-
01
Scan
No app, no signup. A shelf tag reads “not sure which bottle? Scan me” — the shopper scans and AICap opens instantly in their browser.
-
02
Ask
Type or talk, exactly as they'd text a friend who knows the category — occasion, food, taste, budget, who it's for. English, Mandarin, Japanese, or Korean.
-
03
Discover
AICap recommends 2–4 products from that store's real shelf, explains what makes each one fit, and shows exactly where to find it.
-
04
Decide
“Less sweet,” “cheaper,” “compare these two” — the session adapts until the shopper is confident, then they walk straight to the bottle.
Example interaction
Shopper
Dinner tonight — spicy Korean BBQ, something smooth.
AICap
Three that fit: a chilled soju for the table, a sake with enough body to stand up to grilled meat, and one adventurous pick — umeshu, if you want something different.
Shopper
The soju looks a little sweet — anything drier?
AICap
Swapping it for a dry, crisp soju instead — same table-friendly style, less sugar. It's on the third shelf down, right where the sake ends.
Grounded in a real AICap target-shopper scenario — to be replaced with a transcript from live pilot sessions.
Full interaction demo
Greeting → recommendation → close, end to end. Can be re-cut from the Home hero video.
AICap runs on the shopper's own phone — there's no camera and no store-side hardware watching the aisle. Sessions aren't tied to an identity. More on how data is handled below.
Interface
Simple enough for a first-time shopper.
Text is quiet and private for a shopper standing in a public aisle; voice is faster when their hands are full. Both are full experiences — neither is a fallback for the other.
Screen recording or UI mockupA shopper's phone browsing recommendation cards, or a voice-session waveform overlay
Technology
Built to get smarter with every interaction.
Catalog-Grounded Reasoning
AICap never recommends a product the store doesn't carry — every suggestion is reasoned against that store's live catalog, a few steps away on the shelf.
Conversational AI
Understands natural, colloquial shopping language by text or voice, across English, Mandarin, Japanese, and Korean — no rigid menus or keyword commands.
Closed-Loop Learning
Every session — what's asked, recommended, and corrected — trains AICap to explain and recommend better, in every store.
System overview
Catalog + Knowledge
The store's live catalog decides what can be recommended; Onki's knowledge base decides how well it's explained
AI Reasoning
Matches shopper intent to real shelf inventory and explains why each pick fits
Conversational Response
Delivered by text or voice, in the shopper's language, with the product's shelf location
Learning Loop
Every session trains better recommendations, and surfaces assortment gaps & brand insights
Integration & customization
A one-click, read-only connection keeps AICap synced to your real catalog — no truck roll, no manual updates. Content and language mix are configured centrally, per store.
- Shopify integration live today, leading POS systems next
- Prices, availability, and promotions update centrally, in real time
- Language mix configured to each store's shopper base
Data & privacy
AICap collects what was asked, what was recommended, and what happened next — never who asked it. There's no camera, no login, and no personal data tied to a session.
Support & reliability
If a catalog arrives with gaps — missing prices, images, or categories — Onki fixes it with the store owner before going live. A wrong answer at the shelf costs more than a short delay does.
See AICap on your shelf.
See Setup & Pricing
What it takes to get live — catalog sync, signage, timeline.
See Setup & Pricing